CN112602248A - Wind energy installation and method for detecting low-frequency oscillations in an electrical supply network - Google Patents
Wind energy installation and method for detecting low-frequency oscillations in an electrical supply network Download PDFInfo
- Publication number
- CN112602248A CN112602248A CN201980055661.5A CN201980055661A CN112602248A CN 112602248 A CN112602248 A CN 112602248A CN 201980055661 A CN201980055661 A CN 201980055661A CN 112602248 A CN112602248 A CN 112602248A
- Authority
- CN
- China
- Prior art keywords
- frequency
- test
- angle
- measurement
- measurement sequence
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 230000010355 oscillation Effects 0.000 title claims abstract description 129
- 238000000034 method Methods 0.000 title claims abstract description 47
- 238000009434 installation Methods 0.000 title claims description 22
- 238000012360 testing method Methods 0.000 claims abstract description 250
- 238000005259 measurement Methods 0.000 claims abstract description 144
- 238000011156 evaluation Methods 0.000 claims description 15
- 230000002596 correlated effect Effects 0.000 abstract description 7
- 230000006870 function Effects 0.000 description 49
- 230000000875 corresponding effect Effects 0.000 description 13
- 238000003775 Density Functional Theory Methods 0.000 description 7
- 238000004364 calculation method Methods 0.000 description 5
- 238000001514 detection method Methods 0.000 description 5
- 238000011835 investigation Methods 0.000 description 5
- 230000001360 synchronised effect Effects 0.000 description 5
- 238000013016 damping Methods 0.000 description 4
- 230000003595 spectral effect Effects 0.000 description 3
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000033001 locomotion Effects 0.000 description 2
- 101100278475 Saccharomyces cerevisiae (strain ATCC 204508 / S288c) REV3 gene Proteins 0.000 description 1
- 238000009825 accumulation Methods 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 230000005684 electric field Effects 0.000 description 1
- 230000005284 excitation Effects 0.000 description 1
- 230000004807 localization Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000003534 oscillatory effect Effects 0.000 description 1
- 230000010363 phase shift Effects 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 230000000087 stabilizing effect Effects 0.000 description 1
- 238000010561 standard procedure Methods 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
- 238000011144 upstream manufacturing Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R19/00—Arrangements for measuring currents or voltages or for indicating presence or sign thereof
- G01R19/25—Arrangements for measuring currents or voltages or for indicating presence or sign thereof using digital measurement techniques
- G01R19/2513—Arrangements for monitoring electric power systems, e.g. power lines or loads; Logging
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R23/00—Arrangements for measuring frequencies; Arrangements for analysing frequency spectra
- G01R23/16—Spectrum analysis; Fourier analysis
- G01R23/177—Analysis of very low frequencies
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/001—Methods to deal with contingencies, e.g. abnormalities, faults or failures
- H02J3/0012—Contingency detection
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/24—Arrangements for preventing or reducing oscillations of power in networks
- H02J3/241—The oscillation concerning frequency
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/38—Arrangements for parallely feeding a single network by two or more generators, converters or transformers
- H02J3/381—Dispersed generators
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J2300/00—Systems for supplying or distributing electric power characterised by decentralized, dispersed, or local generation
- H02J2300/20—The dispersed energy generation being of renewable origin
- H02J2300/28—The renewable source being wind energy
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E10/00—Energy generation through renewable energy sources
- Y02E10/70—Wind energy
- Y02E10/76—Power conversion electric or electronic aspects
Landscapes
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Measuring Frequencies, Analyzing Spectra (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
- Testing Electric Properties And Detecting Electric Faults (AREA)
Abstract
The invention relates to a method for detecting low-frequency oscillations, in particular subsynchronous resonances, in an electrical power supply network having a network voltage with a nominal network frequency, comprising the following steps: recording at least one measurement sequence of a grid variable, in particular a grid voltage, a feed current or a grid frequency, having a plurality of measurement points, over a measurement period of time in order to perform a frequency analysis; multiplying the measurement sequence with a simultaneously correlated sinusoidal test function over the same measurement time period; wherein the test function is characterized by a test frequency and a test angle as a phase angle; and for each measurement point, multiplying the measurement sequence with a test function to obtain a test product for each measurement point, the test products being added to a product sum taking into account their sign; and evaluating according to the product sum: whether the measurement sequence has a low frequency oscillation having: a frequency in the range of test frequencies, and a phase angle in the range of test angles.
Description
Technical Field
The invention relates to a method for detecting low-frequency oscillations, in particular subsynchronous resonances, in an electrical power supply network. The invention also relates to a wind energy system, i.e. a wind energy installation or a wind farm, for detecting low-frequency oscillations, in particular subsynchronous resonances, in a power supply network.
Background
Many power supply networks increasingly have renewable energy generators, in particular wind energy installations or wind farms. Their share in the power supply network increases to give rise to: it is increasingly important to use wind energy installations and wind farms to support the supply grid, or at least to consider it together.
A problem that can occur in power supply networks, which can also be referred to simply as power networks, is the oscillation, i.e. the oscillation of the energy system, which can also be referred to as "power system oscillation" (PSO). The reasons for this can be very diverse and an intuitive and simple example is that two directly coupled synchronous generators of a conventional power plant oscillate with each other, which synchronous generators can be fed for example more than 100 km from each other.
However, it is also conceivable that a single synchronous generator directly coupled to the supply grid is already excited to oscillate at its natural frequency due to local excitation, for example, sudden changes in the power consumption of the connected consumers. Conventional power supply networks usually deal with this problem by correspondingly stabilizing the synchronous generator which feeds directly into the power supply network. The high inertia of such synchronous generators, together with the damping properties caused physically and/or by the design of the respective generator, generally prevents such oscillations from occurring too strongly in conventional power grids.
However, renewable generators, in particular wind energy installations or wind farms, do not have this property by nature. In particular, they do not have a physically induced characteristic that can counteract such low-frequency oscillations or can avoid them from the outset.
Instead of this, modern wind energy installations or wind farms are currently fed into the supply grid by means of frequency converters using a so-called full power converter design (vollumrichtungrkozept). The entire fed power is thus fed into the supply grid via one or more inverters in accordance with the exact preset value. These preset values relate in particular to the amplitude, frequency and phase of the fed current and can be preset via the process computer. In this case, there is little room for physically induced reactions or for regulating the current fed in.
In order to still be able to react to phenomena in the power supply network, in particular to be able to react to low-frequency oscillations, so-called PSOs, it is therefore necessary to first detect such oscillations ideally in terms of frequency, phase and amplitude. On this basis, the desired reaction measure can then be calculated in the process computer in order to subsequently implement it by means of the inverter.
But possible countermeasures may even worsen the current situation if such low frequency oscillations cannot be detected with sufficient accuracy. Such detection of low-frequency oscillations can be difficult here, since they are first superimposed on the relatively low amplitude of the grid frequency, i.e. the 50Hz or 60Hz voltage signal in the supply grid. In particular when measuring the voltage in the supply network, this can cause the expected occurrence of disturbances and/or noise. Furthermore, such low-frequency oscillations regularly fluctuate. In particular, depending on the stimulus, the low-frequency oscillations occur to a weak or strong extent or even not at all.
Despite the measurement problems, it is desirable to detect as quickly as possible. This again hampers long-term analysis.
A further problem is that such low frequency oscillations may be in the frequency range of 0.05Hz, or even lower, up to a frequency value which can be slightly lower than the grid frequency, i.e. up to an order of magnitude of 50Hz to 60Hz, or even slightly higher. For purely physical reasons, the detection of sinusoidal oscillations requires measurements to be made over the duration of at least one half-cycle of the oscillation. In a large frequency spectrum, a measurement duration at least within a half-period of the oscillation at the lowest frequency expected is therefore required for its detection.
The german patent and trademark office retrieves the following prior art in the priority application for this application: DE 3733555 a1 and US 4,031,462 a.
Disclosure of Invention
The invention is therefore based on the object of being able to solve at least one of the problems mentioned above. In particular, a solution should be proposed which enables the detection of low-frequency oscillations as fast as possible, while it is also able to detect very low-frequency oscillations. At least one alternative solution should be proposed with respect to the solutions known up to now.
According to the invention, a method according to claim 1 is proposed. The method is used for detecting low-frequency oscillations in an electrical supply network, in particular for detecting subsynchronous resonances in an electrical supply network. In this case, a supply network with a network voltage having a nominal network frequency is assumed, wherein the low-frequency oscillations to be detected preferably have a lower frequency than the nominal network frequency. In particular, frequencies lower than the nominal grid frequency are referred to and regarded as low-frequency oscillations. Preferably, a frequency of low frequency oscillation is assumed, which is less than half the grid frequency.
In particular, the low frequency oscillations can have values of 1Hz and lower. However, the low-frequency oscillations can also reach values of up to five times the nominal grid frequency. Oscillations having a frequency of at most a value of five times the nominal grid frequency, preferably at most a frequency corresponding to the nominal grid frequency, are referred to herein as low frequency oscillations. In particular, the low-frequency oscillations do not have a frequency which corresponds to a multiple of the nominal grid frequency. It is to be noted that the investigation and consideration of low-frequency oscillations is used in particular to investigate or ensure the system stability of the power supply network. This is distinguished from the evaluation of the grid quality or the signal quality of the voltage signals in the power supply grid, in which harmonics, in particular integer harmonics, are important.
The method comprises the following steps: at least one measurement sequence having a plurality of measurement points is recorded over a measurement period in order to perform a frequency analysis on the basis thereof. The measurement sequence therefore has a plurality of measurement points or measurement values, i.e. the measurement points or measurement values are distributed over the measurement time period or are recorded in a distributed manner over the measurement time period.
In particular, the grid voltage, the feed current fed into the power supply grid or the grid frequency is recorded as a grid variable.
For this measurement sequence, it is then proposed to multiply the measurement sequence with a sinusoidal test function which is simultaneously correlated over the same measurement time period. Such a sinusoidal test function can be provided as a function on the process computer. It is proposed that the test function is characterized by a test frequency and a test angle as phase angle. The amplitude can also be preset, but in practical implementations it is usually considered to normalize the amplitude, e.g. to a value of 1. The peak value of the sinusoidal deformation quantity can thus for example take the value 1, wherein the scaling can be known in the process computer.
The measurement sequence is multiplied by a test function such that for each measurement point, the measurement sequence is multiplied by the test function to obtain a test product for each measurement point. The measured values of the measurement sequence at the measurement points are therefore multiplied by the corresponding function values of the test function, and this is repeated for each measurement point. The measurement function is accordingly preset for a time duration which corresponds to the measurement time period or within which the measurement function is determined. The multiplication is then carried out separately for each time point of the measurement time period at which the measured value is present. The measured value is then multiplied by the function value of the measurement function at the same point in time.
The test products are now summed to a product sum taking into account their signs. The negative test product is therefore subtracted by the absolute value. I.e. each test product can have a negative sign if the measurement value concerned or the function value concerned is negative.
Then, the following is evaluated according to the product sum: whether the measurement sequence has a low-frequency oscillation with a frequency in the range of test frequencies and a phase angle in the range of test angles.
This is based in particular on the following considerations. If the test function and the measurement sequence are identical in terms of frequency and phase, each test product has a positive value and accordingly the product sum is likewise positive and has a larger amplitude. The measurement sequence then reproduces a function corresponding to the test function.
If the measurement sequence and the test function are identical but phase-shifted by 90 ° from one another, the test product will result in a sine-shaped function (which in addition has twice the frequency for the test function), to be precise without a direct current component. Then, the product sum will be zero at least when the investigation period corresponds to an integer multiple of the period of the investigated signal. Furthermore, when the current and voltage are offset from each other by 90 °, this phenomenon is known as reactive power to electrical engineers. The first mentioned example without phase shift corresponds to the case with active power only. This will produce a sinusoidal signal of double frequency that moves so much around the horizontal axis that the minima of the sinusoidal function just touch the horizontal axis. This movement can also be understood or referred to as a direct current component. The dc component will be high, i.e. highest.
However, deviations in frequency between the measurement sequence and the test function may also cause different dc components. The product sum, which can be interpreted as a direct current component or which can represent a direct current component, can therefore also be used as a measure of the correlation between the measurement sequence and the test function. If the measurement sequence and the test function are completely uncorrelated, the direct current component becomes zero, at least theoretically for an infinite measurement period.
But even in practical implementations the product sum is at least relatively small if the measurement sequence and the test function are not correlated. If the measurement sequence and the test function are well correlated and the phase angles of the measurement sequence and the test function also match one another, a high direct current component or a large product sum results. From which the presence of low frequency oscillations can be deduced in terms of frequency and phase. The low frequency oscillation then has the frequency of the test function and a phase angle corresponding to the test angle of the test function.
Preferably, the amplitude of the low-frequency oscillations is detected, in particular as a product sum, when low-frequency oscillations are identified. In particular, it has been recognized that the amplitude can also be determined via a method for identifying low-frequency oscillations from the frequency and the phase.
Preferably, the method is carried out such that, with a change in the test frequency and a change in the test angle, the multiplication of the measurement sequence by the test function and the addition of the test products to a product sum are repeated in order to obtain a plurality of product sums here. In principle, it is also advantageous to change only the test frequency or the test angle, but it is often advantageous to change the test frequency and the test angle as long as neither the frequency to be identified nor its phase of the low-frequency signal is known. The result of the product sum can also be shown as a curved plane if both values change. The product sum is then shown in terms of test frequency and test angle. Thus, a local maximum will be generated in case of test frequencies and test angles where there is an optimal agreement between the measurement sequence and the test function.
It is therefore proposed that the evaluation is performed on the basis of the plurality of product sums thus obtained: whether the measurement sequence has low frequency oscillations. All the product sums obtained here can therefore be compared and from this the presence of low-frequency oscillations can be identified, as long as they have a frequency and a phase which are respectively located in the vicinity of the test frequency or the test angle.
In particular, therefore, in the case of a product sum having the greatest amplitude relative to the remaining product sums, a low-frequency oscillation is assumed which has the frequency and phase of the relevant test frequency and of the relevant test angle. The above-mentioned three-dimensional views do not necessarily need to be selected for the evaluation of all product sums. It is also possible to consider: only the maximum value of the product-sum is searched, or in the simplest case the maximum product-sum.
It is therefore proposed in particular that, with a change in the test frequency and a change in the test angle, the measurement sequence is multiplied by the test function and the test products are added to a product sum in such a way that the product sum is recorded for each test pair formed by the test frequency value and the value of the test angle. This is done in particular in such a way that the recorded product sum can be displayed as a curved area in three-dimensional space in dependence on the test frequency value and in dependence on the test angle, wherein the product sum does not necessarily have to be displayed.
In this case, it is proposed, in particular for the purpose of evaluation, that in the case of a product sum having the greatest amplitude relative to the remaining product sums, a low-frequency oscillation is assumed which has the frequency and phase of the relevant test frequency and of the relevant test angle. In principle, of course, also: the interpolation is performed if a number of adjacent product sums form a maximum value, i.e. are of the same size or almost the same size.
This also enables, in particular, the evaluation to be programmed fully automatically. Basically only the frequency range to be tested needs to be preset. The step size by which the test frequency is changed can also be preset if necessary.
The test angle is preferably fully tested or fully changed from 0 ° to 360 °. If necessary, the test angle is completely tested from 0 ° to 180 °, and when searching for the maximum product sum, the product sum that is the largest in absolute value can be searched for.
According to one embodiment, it is provided that, in order to assume low-frequency oscillations of the product sum, only or additionally: whether the product sum reaches at least a predetermined test magnitude. This assumption is in particular that, i.e. based on actual implementation only, in any case allows the maximum product sum to be expected if the test frequency and the test angle are changed completely as described above. From which it can then not be deduced: in practice, low-frequency oscillations or associated low-frequency oscillations are present, since even very small maxima may be caused by measurement inaccuracies or other influences.
And (3) testing: the product sum at least reaches a predetermined test amplitude, if necessary, also enables a plurality of low-frequency oscillations to be identified, if they should be present. In particular, for intuitive interpretation of curved surface selection based on product sum: local maxima are also found in the curved surface by testing whether a predetermined test amplitude is reached.
Of course, in the case of a plurality of product sums which are adjacent to one another or at least close to one another in terms of frequency and phase angle, it is generally not possible to assume that a plurality of low-frequency oscillations are present. The accumulation of such large product sums then points to a maximum value within which the maximum product sum is preferably selected as the maximum product sum whose test frequency and test angle assume the frequency and phase of the identified low-frequency oscillation.
According to one specific embodiment, it is provided that, in a first test cycle, the multiplication of the measurement sequence by the test function and the summation of the test products are repeated with a changing test frequency. The test frequency is varied in a first frequency range in order to detect low-frequency oscillations having an oscillation frequency. The oscillation frequency is first detected with a first accuracy. In a second test cycle, the multiplication of the measurement sequence by the test function and the summation of the test products are repeated with changing test frequencies. Then, in the case of this repetition in the second test cycle, the test frequency is changed within the second frequency range. For this purpose, the second frequency range is selected in accordance with the oscillation frequency identified in the first test cycle, in order to detect the oscillation frequency with a higher accuracy than in the first cycle.
This is based in particular on the following idea: testing many test frequencies and many test angles completely can result in a large number of tests. It is therefore proposed that upstream a first test cycle is entered, which changes only the test frequency, if necessary in a first frequency step, which is greater than a second, later frequency step, by means of which the test frequency is changed in a second test cycle. By means of this first test cycle, a first overshoot in absolute value should occur, at least in the vicinity of the frequency of the low-frequency oscillation.
And then more accurately tested in this first overshoot zone. For this purpose, a second frequency range is predefined, which is in particular smaller than the first frequency range, in particular lies in the first frequency range. The test angle can then also be changed in the second test cycle. This enables a high degree of accuracy to be achieved without having to carry out this high degree of accuracy over the entire theoretical range, which is produced by the first frequency range and the entire 360 ° range of the test angle.
In this case, in the first test cycle, the multiplication of the measurement sequence by the test function and the summation of the test products are preferably repeated, in addition to the change in the test angle. Thus changing the test frequency and test angle. The test angle is changed in this case in a first angle range, in particular in the range from 0 ° to 360 °, but the angle can also preferably be limited to 0 ° to 180 ° or another range including 180 °. This is to additionally identify low-frequency oscillations by means of the phase angle of the oscillations. Furthermore, if the phase angle and the test angle differ greatly from one another, the product sum can have a value which is low in absolute value despite a well-matched frequency between the low-frequency oscillation and the test function. Then low frequency oscillations may be ignored.
For this purpose, it is also proposed to change the test angle in a first angle step, i.e., for example in 5 ° steps. In a second test cycle, the multiplication of the measurement sequence by the test function and the summation of the test products are additionally repeated with a change in the test angle, the test angle changing within a second angle range. The change is preferably carried out in a second angle step which is smaller than the first angle step, for example in a second angle step of 1 °. In particular, the second angular range is selected in dependence on the phase angle of the oscillation identified in the first test cycle, in order to detect the phase angle of the oscillation with a higher accuracy than in the first cycle.
It is therefore proposed that the test frequency and the test angle are changed in a first test cycle, and that a first coarse identification of at least one low-frequency oscillation is carried out, i.e. an approximation of the frequency of the low-frequency oscillation and an approximation of the phase angle of the low-frequency oscillation are determined. The search can then be improved in a range around the frequency and the angle by: where the test frequency and test angle are fully tested over a smaller frequency range and also over a smaller angular range.
According to one embodiment, it is provided that in a first test cycle the test frequency is changed in a larger frequency step than in a second test cycle, and that in addition or alternatively in the first test cycle the test angle is changed in a larger angle step than in the second test cycle. In this way, a large frequency range can be completely tested in the first test cycle with acceptable effort. It is also possible to fully test a large angular range, i.e. in particular a complete angular range from 0 to 360 °, at least from 0 to 180 °. This enables a first localization of the low-frequency oscillations to be carried out at an acceptable cost. Then, in the second test cycle, an accurate search with a higher accuracy is only required for a smaller range, i.e. a smaller frequency range and a smaller angular range.
According to a further embodiment, a plurality of measurement sequences of the grid variable are recorded. A plurality of measurement sequences, in particular the network voltage, the feed current or the network frequency, is thus recorded. Each measurement sequence is provided for evaluating a frequency range. For this purpose, a measurement time period is selected for each measurement sequence depending on the frequency range to be analyzed. Different frequency ranges are therefore investigated and for this purpose corresponding measurement sequences are recorded in each case. In particular, it is proposed for this purpose that the measurement time period for analyzing the frequency range of the small frequencies is selected to be long, i.e. so long that the lowest frequency of the frequency range can still be detected. Accordingly, a shorter measurement time period can be provided for a higher frequency range. Furthermore, it is preferably provided that, in a correspondingly long measuring period, the measuring points in time are further apart than in a shorter measuring period.
It is therefore proposed to divide the frequency range to be investigated overall, for example 0.1Hz or less, up to 50Hz or even 250Hz, into at least two measurement ranges, in particular into a frequency range of low frequencies and a frequency range of higher frequencies. It is therefore divided into a first and a second frequency range, possibly into further frequency ranges. For each frequency range to be analyzed, i.e. for example for a frequency range of low frequencies and a frequency range of higher frequencies, a measurement sequence is recorded. For the example mentioned, two measurement sequences are therefore recorded.
Then, with changing the test frequency and additionally or alternatively with changing the test angle, for each measurement sequence, the multiplication of the measurement sequence with the test function and the addition of the test products to a sum are repeated in order to obtain a plurality of product sums for each measurement sequence. Furthermore, it is proposed that for each measurement sequence the product sum of the associated measurement sequences is evaluated separately in order to detect low-frequency oscillations.
In particular based on the following recognition: in order to study the frequency range of low frequencies on the one hand and the frequency range of higher frequencies on the other hand, the measurement times should be taken into account reasonably on the basis of different measurement times, i.e. measurement time periods. On the other hand, for low frequency oscillations with relatively high frequencies, such long measurement periods may be too long to identify the low frequency oscillations in time.
For example, the first frequency range, which can also be referred to as the low frequency range, can be from 0.02Hz up to 2 Hz. In order to be able to detect low-frequency oscillations with 0.02Hz, at least a reciprocal value of the measurement period should be used, i.e. a measurement period of 50 seconds. Then, a second frequency range, which may also be referred to as a higher frequency range, can be reached, for example, from 2Hz to 250 Hz. Here, a measurement duration of 0.2 seconds is sufficient to detect even the lowest frequency of 2 Hz. In this case, oscillations in such a second or higher frequency range may increase significantly over a test period of 50 seconds of the first frequency range, if necessary, in extreme cases up to a resonance disaster or at least to the point where the amplitude of the low-frequency oscillations becomes so great that first damage may occur or a first shutdown process may be initiated.
In order to cope with this problem, it is proposed here to divide the frequency range into at least two frequency ranges and to carry out the evaluation in particular also independently of one another in time.
According to one embodiment, it is provided that the multiplication of the measurement sequence by the test function and the addition of the test products to the product sum are repeated with a specific change to the test frequency. I.e. the test frequency is changed in frequency steps in at least one frequency range having an upper and a lower frequency value, and the frequency steps are set in accordance with the frequency range. This is particularly so that the frequency step is smaller than the lower frequency limit value, in particular smaller than 10% of the lower frequency limit value. In addition or alternatively, it is provided that the frequency step is smaller than a predetermined percentage value, in particular that the frequency step is set to be smaller than 1%, in particular smaller than 0.2%, of the upper frequency limit value.
Preferably, the test frequency is changed in a plurality of frequency ranges, and the frequency steps of different frequency ranges are set differently from each other. Preferably, each frequency step is set to a predetermined percentage fraction greater than the respective lower frequency limit value of the associated frequency range.
The test frequency is thus changed according to the respective frequency step in the respective frequency range. The frequency step is set here as a function of the frequency range and is adapted here in particular to the lowest frequency value of the respective frequency range. For this purpose, a percentage value can be provided with respect to the lower frequency limit value. However, the frequency step can also be adapted to the upper frequency value, but the frequency step is also selected to be relatively much smaller with respect to the upper frequency value of the associated frequency range. In particular, it is thereby achieved that the variation of the test frequency is clearly predetermined and is also selected differently for different frequency ranges. The test complexity, i.e. the complexity due to the variations, can thus be adapted to the respective frequency range. The presets also enable automated test routines.
According to the invention, a wind energy system is also proposed. Such a wind energy system can be a wind energy installation or a wind farm with a plurality of wind energy installations. The wind energy system is fed into the supply grid according to the regulations. The wind energy system is configured to detect low-frequency oscillations, in particular subsynchronous resonances, in the supply grid. The power supply network has a network voltage with a nominal network frequency. The wind energy system comprises:
recording means for recording at least one measurement sequence of grid variables, in particular grid voltage, feed current or grid frequency, over a measurement period for performing a frequency analysis, wherein the measurement sequence has a plurality of measurement points,
a multiplication unit for multiplying the measurement sequence with a sinusoidal test function which is simultaneously correlated over the same measurement time period, wherein
-the test function is characterized by a test frequency and a test angle as a phase angle, and
-multiplying the measurement sequence for each measurement point with a test function to obtain a test product for each measurement point,
-an adding unit for adding the test products to a product sum taking into account their signs, and
-evaluation means for evaluating, from the product-sum: whether the measurement sequence has a low frequency oscillation having:
-a frequency in the test frequency range, and
-a phase angle within the test angle range.
Such a wind energy system can therefore be fed into the supply grid and preferably also undertakes a supporting task for supporting the supply grid. Such a support task can be particularly necessary or at least advantageous if the decentralized generator and the wind energy system feed a large portion into the power supply network or into the relevant section of the power supply network. Various support tasks may occur, one of which can be to react to low frequency oscillations. Preferably, however, such low-frequency oscillations are detected first, as precisely as possible in terms of frequency and phase, and if appropriate also in terms of amplitude. And then able to react to it.
Preferably, the wind energy system is configured to carry out at least one method according to the above-described embodiment. In particular, the wind energy system has a process computer for this purpose, which is configured to carry out such a method. In particular, the method is implemented on a process computer for this purpose. The implementation of the method can here comprise: the recording of the measurement point or the measurement sequence or measurement sequences is carried out by: the process computer receives the respective values as measurement points or values and/or the process computer controls the recording means in order to record at least one measurement sequence therefrom.
In particular, the recording means can be a sensor, which measures, for example, a voltage or a current. The multiplication units can likewise be implemented in the same or different process computers. The same applies to the addition units, although these units can also constitute different equipment units. The evaluation device can likewise be implemented in a process computer or in the same process computer or else be provided as a separate element.
Drawings
In the following, the invention is now explained in detail, by way of example, with reference to the accompanying drawings.
Fig. 1 shows a perspective view of a wind energy installation.
Fig. 2 shows a schematic diagram of a wind farm.
Fig. 3 shows a flow pattern for recording a plurality of product sums in the case of changing the test series and changing the test angle.
Fig. 4 shows a flow chart for evaluating a plurality of product sums recorded according to the flow chart of fig. 3.
Fig. 5 shows a 3D graph of product sums according to varying test systems and varying test angles.
FIG. 6 schematically shows the structure of a wind energy system for identifying low frequency oscillations.
Detailed description of the preferred embodiments
Fig. 1 shows a wind energy plant 100 with a tower 102 and a nacelle 104. A rotor 106 having a fairing 110 and three rotor blades 108 is disposed on nacelle 104. During operation, rotor 106 is set into rotational motion by the wind, which in turn drives a generator in nacelle 104.
Fig. 2 shows a wind farm 112 with exemplary three wind energy installations 100, which can be identical or different. Thus, the three wind energy installations 100 represent essentially any number of wind energy installations in the wind farm 112. The wind energy installation 100 is supplied with its power, i.e. in particular the generated current, via a power grid 114 of a power plant. The respectively generated currents or powers of the individual wind energy installations 100 are summed and a transformer 116 is usually provided, which converts the voltage in the wind farm up to a voltage level in order to then feed it into a supply grid 120 at a feed point 118, which is usually also referred to as PCC. Fig. 2 is only a simplified view of the wind farm 112, although of course a control device is present, said fig. 2 for example not showing a control device. The electric field network 114 can also be designed differently, for example, in that: for example, a transformer is also present at the output of each wind energy installation 100, just to name a further example.
The wind power installation according to fig. 1 and the wind farm according to fig. 2 can each form a wind energy system.
Fig. 3 shows a flow chart 300 for recording multiple product sums. In a start block 302 the signal to be investigated is recorded and further initialisation takes place. The signal to be investigated can be a recorded time signal which is sampled uniformly in the flow chart 300 in time steps Δ t for the investigation. The signal to be investigated can also already be present in this sampled form, but the time step is advantageously selected here in order to thereby also determine the number of values to be investigated overall.
The signal y (t) to be investigated is therefore recorded or taken into account during the measurement period, anAnd the measuring time period can be from t-0 to t-tend. Thus, the measurement duration and thus the width of the measurement period pass through tendTo be determined. For time t therefore applies:
t＝0，1·Δt，2·Δt，...，tend
likewise, in start block 302, the starting frequency f can be determinedstartTo the end frequency fendThe frequency range to be investigated. The step size Δ f of the frequency study can be based on the starting frequency fstartAnd an end frequency fendAnd a desired number of frequency steps n, determined according to the following equation:
Δf＝(fend-fstart)/n
the step size of the phase angle study can likewise be determined as a function of the desired number m of angle steps
Determined according to the following formula:
these values, in particular the number of frequency steps n and angle steps m and the time step Δ t, can in principle be chosen arbitrarily, but it is proposed to make a trade-off between accuracy and computational effort in the selection.
In a first initialization block 304, a control variable i for the outer loop 306 is initialized. The outer loop 306 runs as often as the number of frequency steps n and is accordingly tested in a first repeat query block 308 following the first increment block 310.
Inside this outer loop 306 is a second initialization block 312 in which the control variable j for the inner loop 314 is initialized. The control variable is run in terms of angular steps m, which is queried in a second repeat query block 316 following a second increment block 318.
Finally, a calculation block 320 is provided, which runs (n × m) times accordingly. Computing parameters in each runFrequency of examination frefI.e. the frequencies, for which the sum products are calculated, respectively. Calculating the reference frequency f according to the following formularef：
fref＝fstart+i·(fend-fstart)/n
At the same time, the corresponding reference angle theta is calculatedrefI.e. according to the following calculation:
θref＝(j·2·π)/m
finally, the values calculated based on these are then used, i.e. for the reference frequency frefAnd a reference angle thetarefThe product-sum is calculated. As explained above, the product sum can also be understood as a direct current component, so that the product sum is referred to herein as DCprod. Thus, the product-sum is calculated for the respective run i of the outer loop and the respective run j of the inner loop according to the following formula:
DCprod(i，j)＝Summe{y(t)·sin(2·π·fref·t+θref)}·Δt/tend
the signal y (t) to be investigated as a measurement sequence is therefore compared with the sine function sin (2. pi. f)ref·t+θref) Multiplied and a sum is formed in respect thereof. Thus, here too, a product is formed for each time point and the products are summed. This can be performed, for example, by the third innermost loop, in order to illustrate this intuitively, in which the time t increases from 0 to tendI.e. increased by a time step at. Furthermore, by multiplying by the time step Δ t and dividing by the end time tendThe result can also be normalized, i.e. such that the product sum is DCprodIn principle independent of the time step deltat. The product sum is therefore in principle independent of the number of products that are added and summed with respect to its absolute value.
After the inner loop 314 has been run m times and the outer loop 306 has been run n times, then there are n x m separate product sums DCprod(I, j) that the product sum can be saved in the corresponding field and then studied for further evaluation. To this end, the flow chart 300The result is passed to the flow chart 400 of fig. 4, which is represented in the flow chart 300 by block 400.
Accordingly, fig. 4 shows this block 400, i.e. the flow chart 400, and this is based on the flow chart 300 of fig. 3, which is represented by the way the first block is referred to as flow chart 300.
In a maximum block 402, the product sum with the maximum value is found from all product sums calculated in the calculation block 320, i.e. taking into account the sign. If, for the sake of simplicity or reduced complexity, the test angle does not change within 360 °, but only within 180 °, it is preferably also considered to change the test angle only within 90 °, where the value with the largest absolute value can also be searched for.
A search is made for all product sums stored in particular in the fields, i.e. according to the operation of the outer loop 306 and the inner loop 314. That is, the loops are run with the external control variable i and the internal control variable j, and these two control variables are then also used here to identify the largest product sum, for example in the data field. Accordingly, the assignment of these two control variables is carried out in the recognition block 404, i.e. the control variables i and j are recognized as the selected external control variable i, on the basis of which the greatest product sum is found in the maximum block 402 for the control variablesMaxDCAnd selected internal control variable jMaxDC。
The maximum of the product-sum identified in block 402 belongs to the two selected controlled variables, namely the selected external controlled variable and the selected internal controlled variable iMaxDCOr jMaxDCAnd the reference frequency and the reference angle belong to the maximum value. The respective reference frequency and the respective reference angle can be calculated from the respective external and internal control variables i, j. For this frequency or this angle it is assumed that: this is the corresponding frequency of the low-frequency oscillation or the corresponding angle of the low-frequency oscillation, so that the associated reference frequency is referred to as the frequency f of the low-frequency oscillationPSOAnd the selected angle is referred to as the angle theta of the low frequency oscillationPSO. These two values can be calculated according to the following equation:
fPSO＝fstart+i·(fend-fstart)/n
θPSO＝(j·2·π)/m
and if the corresponding control variable i is to be selectedMaxDCOr jMaxDCFor the corresponding external or internal control variable i, j, the frequency f of the low-frequency oscillation is then calculatedPSOAnd angle theta of low frequency oscillationPSO. In this formula, the angle θ of the low frequency oscillationPSOIs marked in radians rather than degrees.
The amplitude a of the low frequency oscillations can also be calculated in the calculation block 406PSO. I.e. calculated according to the following equation:
APSO＝MaxDC/(Summe{sin(2·π·fPSO·t+θPSO)·sin(2·π·fPSO·t+θPSO)}·Δt/tend)
thus, the amplitude of the low frequency oscillations is generated as follows: the detected maximum product sum is divided by the corresponding product sum of the reference signal multiplied by the reference signal over the time range of the study. The product-sum of the reference signals is thus determined when multiplied by itself, which leads to a possible maximum, since such a reference function is most strongly correlated with itself. Thus preserving the factor of the product sum with a lesser degree of correlation between the signal under consideration and the reference signal. In this case, the amplitude APSOAgain a normalization variable.
The result can thus be output in an output block 408 and further used.
Fig. 5 shows diagrammatically in a three-dimensional view 500 the entirety of all product sums which are plotted in a computation block 320 as curved planes 502 according to a varying reference frequency frefAnd a varying reference angle thetarefAnd (4) calculating. The reference frequency frefCan also be synonymously referred to as test frequency, and the reference angle thetarefCan also be synonymously referred to as a test angle.
The signal to be investigated is exemplarily selected to have a phase angle of 90 ° (θ)PSOThe oscillation frequency in the case of 90 ° is 8.25Hz (f)PSO8.25 Hz). For this purpose, the reference angle or test angle is changed completely from 0 ° to 360 °, and the reference frequency or test frequency is changed completely from 0 to 25 Hz. It can be recognized that the value of the product sum plotted in the curved face 502 is substantially 0 for frequencies that deviate greatly from this oscillation frequency of 8.25 Hz. Around an oscillation frequency of 8.25Hz, the amplitude increases in an oscillating manner towards the oscillation frequency. It can also be seen that the reference angle or the test angle also plays an important role. The absolute amplitude of the product sum is then also the greatest in the case of the oscillation frequency and the phase angle of the low-frequency oscillation, and accordingly the frequency f of the low-frequency oscillation can be read from the table or the value field of the product sumPSOAnd angle theta of low frequency oscillationPSO。
The flow chart of fig. 3 and the flow chart of indirect map 4 and the chart of fig. 5 relate to the following case: the test or reference frequency and the test or reference angle, respectively, are changed only once, although with a number of values, respectively, but the entire sequence of the outer and inner loop is not repeated with new values, in particular according to the flow diagram 300. This view, in particular of fig. 3, is used for illustration purposes and preferably the entire sequence according to the two flowcharts 300 and 400 is repeated with focused values, in particular for the frequency to be investigated, i.e. for the frequency range to be investigated and also for new values of the angle range to be investigated. For this purpose, in the starting block 302, the frequency f for the low-frequency oscillation is provided in the output block 408, in particular on the basis of the first operationPSOAnd phase angle theta of low frequency oscillationPSOAccordingly, a new value around the roughly identified maximum value is determined.
Fig. 6 shows a wind energy system 600, which is illustrated symbolically by a single wind energy installation, but can also have a plurality of wind energy installations. The wind energy installation is configured to detect subsynchronous resonances in a supply grid 602 to which the wind energy system 600 is fed.
A recording means 604 is provided for recording at least one measurement sequence of the network variable, which recording means are able to detect the network voltage, the feed current or the network frequency. The measurement sequence thus detected can be sent to a multiplication unit 606, which can perform a multiplication with a test function sin (t). The test function sin (t) is only mentioned symbolically in this connection and, as also mentioned above, is more complex than such a sine function and can be varied at least in respect of some input variables.
The result of the multiplication unit 606 is passed to an addition unit 608, where the test products generated in the multiplication unit 606 are added to a product sum. The product sum is thus the result of the addition unit 608 and the result is provided to the evaluation means 610. The evaluation device here searches for the maximum of all product sums which it obtains from the multiplication unit 606. For this purpose, a storage device 612 for recording data fields can be provided, which is shown here as part of the evaluation device 610. If low-frequency oscillations are found, the result of the evaluation means is ultimately its oscillation frequency fPSOAnd its phase angle thetaPSO. These values can then be further processed by a further process computer 614, for example, in order to adjust the feed of the wind energy system 600 into the supply grid 602 in such a way that such recognized oscillations are cancelled out. Furthermore, these two values, i.e. the frequency and the phase angle of the low-frequency oscillation, can be fed back to a synthesis block 616, which generates the already described test function, which is symbolically shown as sin (t), or which is adjusted in a further cycle. In particular, the input values are adjusted.
Thus, in particular, consider: identifying low frequency oscillations (PSO/power system oscillations) and their parameters can be a challenge. This is due in particular to: low frequency oscillations typically have very low frequency components. The problems not only lie in: it is indeed recognized that oscillations are present, but that they are also recognized, i.e. in particular, which frequency, which phase angle and which absolute value of the oscillation is present.
Known DFT methods can in principle be used. It has been recognized, however, that such DFT methods can provide information over a wide frequency range, depending on the sampling rate of the signal, which is not necessarily helpful. Furthermore, the DFT method requires a correspondingly long time window in the time domain for finer resolution in the frequency domain. It has been recognized herein that the frequencies expected during low frequency oscillations lie within a limited frequency range, and that this can be exploited so that other effective ways of focusing on this limited frequency range can be helpful. Advantageously, the corresponding method also requires a shorter time window.
The invention is also based on the following recognition: the energy system is an oscillatory system with natural modes below and above the system frequency (50Hz, 60 Hz). When excited, such oscillations can affect system stability if they cannot be sufficiently damped. Here, a new way of detecting so-called Power System Oscillations (PSO) is now proposed. A feasible accurate identification of the frequency, phase angle and absolute value of the Power System Oscillation (PSO) from the signal is to be achieved.
The observation of the Power System Oscillations (PSO) can not only be helpful as an alarm system for operating a wind farm, but this information can also be used as a basis for suitably generating a damping signal by the wind energy installation or wind farm for damping the power system oscillations.
In particular, it is also recognized that observing Power System Oscillations (PSOs), i.e. in particular low frequency oscillations, can also be an important component of an alarm system for the operation of a wind farm. Furthermore, most methods for damping PSO are based on the accurate identification of oscillations from the measurements.
The proposed method enables in particular to identify the PSO (or other type of oscillation) and its most important characteristics (frequency, phase angle and absolute value).
The proposed method is particularly directed to a feasible accurate detection of the frequency, phase angle and absolute value in the measured signal by means of a measurement window that is as short as possible. Frequent limitations of real systems, such as computing power, memory space for measurement data, assumptions about a constant operating point, are taken into account here.
The proposed method is based on the following principle: the signal to be investigated having a frequency frefDirect current component of the product of the sinusoidal reference signal of (a)DC component) only with the signal at frequency frefIs correlated. All other signal components, which therefore do not have the frequency of the reference signal, are averaged out to some extent to be expressed graphically.
The concept on which this is based can be summarized as follows. The signal to be investigated is multiplied with a sinusoidal reference signal. The phase angle of the reference signal is varied in a loop over the entire range (0 to 2 pi or 0 ° to 360 °) by m iterations. Furthermore, the frequency of the reference signal is in another cycle in the frequency range (f) to be investigatedstartTo fend) The inner is changed by n iterations. Thus producing m n products. The frequency and phase angle at which the DC component of the product is highest can be assumed to be the frequency and phase angle of the low frequency oscillation. The absolute value of the low frequency oscillation can also be determined by knowledge of the frequency and phase angle. This flow is basically illustrated diagrammatically in fig. 3 and 4.
If the two cycles 306 and 314 according to fig. 3 are thus run more frequently and in smaller steps, the parameters m and n can be increased to improve for a particular frequency range (f)startTo fend) The accuracy of the method of (1). One possibility for optimizing the computational effort is to implement the proposed method according to fig. 3 and 4 in two stages:
1. the first stage has a coarse resolution ((f)end-fstart) N) and as a result provides the frequency of the PSO (hereinafter referred to as f)PSO1) A coarse estimate of (2). For this purpose, the following values are proposed for the study parameters:
fstart1＝fstart
fend1＝fend
n1according to (f)end－fstart)·tendThe next integer of 2
m1＝36
2. The second stage then investigates a smaller frequency range around the result of the first stage with a finer resolution. For this purpose, the following values are proposed for the study parameters:
fstart2＝fPSO1－1/(tend·2)
fend2＝fPSO1+1/(tend·2)
n2: as high as possible (≧ 2)
m2: as high as possible (≧ 36)
n1 and n2 denote the first or second number of repetitions of the cycle for frequency change.
m1 and m2 denote the first or second number of repetitions of the cycle for phase angle change.
Advantages over FFT and DFT (standard methods):
in FFT and DFT it is possible to recognize only oscillations with a specific frequency, namely: the frequency of which corresponds to an oscillation of an integer multiple of 1/T (T: the length of the investigation time window). Since the frequency of the PSO is an unknown variable, it is highly unlikely that the frequency of the PSO randomly corresponds to an integer multiple of 1/T. Therefore, when applying FFT or DFT, a certain error is always expected when determining the frequency. Errors in determining frequency can prevent the determination of phase angle and absolute value. In contrast to FFT and DFT, the frequency range can be studied arbitrarily fine in the proposed method. In this case, a compromise is only made between accuracy and computational expense. The FFT provides information about a particular spectral line as a result. The number of spectral lines is related to the number of measurement points of the signal to be investigated. It is not feasible to study a part of the spectral lines in the case of FFT. In other words, there is no possibility of performing FFT calculation for a limited frequency range. In contrast, the investigation range for the frequency can be selected arbitrarily in the proposed method. It is also possible to select the computational effort by selecting a study resolution that matches the available computational power.
Claims (11)
1. Method for identifying low-frequency oscillations, in particular subsynchronous resonances, in an electrical power supply network, wherein the electrical power supply network has a grid voltage with a nominal grid frequency, comprising the steps of:
recording at least one measurement sequence of grid variables, in particular grid voltage, feed current or grid frequency, over a measurement period to perform a frequency analysis, the measurement sequence having a plurality of measurement points,
-multiplying the measurement sequence with a sinusoidal test function of simultaneous correlation in the same measurement time period, wherein
-the test function is characterized by a test frequency and a test angle as a phase angle, and
-for each measurement point, multiplying the measurement sequence with the test function to obtain a test product for each measurement point,
-adding the test products to a product sum taking into account the sign of the test products, and
-evaluating from the product sum: whether the measurement sequence has a low frequency oscillation having:
-a frequency in the range of the test frequency, and
-a phase angle in the range of test angles.
2. The method of claim 1,
-repeating multiplying the measurement sequence with the test function and adding the test products to a product sum, in case the test frequency is changed and/or in case the test angle is changed, in order to obtain a plurality of product sums, and
-performing an evaluation on the basis of the plurality of product sums thus obtained: whether the measurement sequence has low-frequency oscillations, in particular, such that
In the case of a product sum having the largest amplitude relative to the remaining product sums, a low-frequency oscillation is assumed, which has the frequency and phase of the associated test frequency and of the associated test angle, and/or
-detecting the amplitude of the low frequency oscillations upon identification thereof, in particular from the product sum.
3. The method according to claim 1 or 2,
-repeating multiplying the measurement sequence with the test function and adding the test products to a product sum, with changing the test frequency and with changing the test angle, such that
-recording a product sum for each test pair formed by a test frequency value and a value of said test angle, wherein in particular
-the recorded product sum is displayable as a curved surface in three-dimensional space in relation to the test frequency value and in relation to the test angle, and wherein
-assuming test angle and test frequency values of a test pair, for which the product sum forms a maximum value with respect to the remaining product sums recorded, as phase angle and frequency of the low frequency oscillation.
4. The method according to any of the preceding claims,
to assume low frequency oscillation of the product-sum, only or additionally tested: whether the product sum reaches at least a predetermined test magnitude.
5. The method according to any of the preceding claims,
-repeating multiplying the measurement sequence with the test function and summing the test products with changing the test frequency in a first test cycle, wherein
-varying the test frequency in a first frequency range in order to identify low frequency oscillations having an oscillation frequency, wherein
-detecting said oscillation frequency with a first accuracy, and
-repeating multiplying the measurement sequence with the test function and summing the test products with changing the test frequency in a second test cycle, wherein
-varying the test frequency in a second frequency range, and
-selecting the second frequency range in dependence on the oscillation frequency identified in the first test cycle, so as to detect the oscillation frequency with a higher accuracy than in the first cycle.
6. The method of claim 5,
-in the first test cycle, in addition to changing the test angle, repeating the multiplication of the measurement sequence with the test function and the summation of the test products, wherein
-changing the test angle in a first angle range, in particular in a range of 0 ° to 360 °, wherein
-changing the test angle in first angle steps, and
-in the second test cycle, the multiplication of the measurement sequence with the test function and the summation of the test products are repeated, additionally with changing the test angle, wherein
-changing the test angle within a second angle range, in particular in a second angle step, which is smaller than the first angle step, and
-selecting the second angular range in dependence on the phase angle of the oscillation identified in the first test cycle, so as to detect the phase angle of the oscillation with a higher accuracy than in the first cycle.
7. The method according to claim 5 or 6,
-in the first test cycle, changing the test frequency in larger frequency steps than in the second test cycle, and/or
-in the first test cycle, changing the test angle in larger angular steps than in the second test cycle.
8. The method according to any of the preceding claims,
recording a plurality of measurement sequences of the grid variable, in particular the grid voltage, the feed current or the grid frequency, wherein
Each measurement sequence is arranged for analyzing a frequency range,
-selecting a measurement time period for each measurement sequence according to the frequency range to be analyzed, and
-repeating, for each measurement sequence, multiplying the measurement sequence with the test function and adding the test products to a test sum, with changing the test frequency and/or with changing the test angle, so as to obtain a plurality of product sums for each measurement sequence, a sum
-evaluating, for each measurement sequence, the product sum of the associated measurement sequences separately to identify low frequency oscillations.
9. The method according to any of the preceding claims,
-repeating multiplying the measurement sequence with the test function and summing the test products to a product sum with changing the test frequency, wherein
-changing the test frequency in frequency steps over at least one frequency range, the frequency range having an upper frequency value and a lower frequency value, and
-setting the frequency step size as a function of the frequency range, in particular such that
The frequency step is smaller than the lower limit frequency value, in particular smaller than 10% of the lower limit frequency value, and/or
-said frequency step is smaller than a predetermined percentage value of said upper frequency limit value, preferably smaller than 1% of said upper frequency limit value, in particular smaller than 0.2% of said upper frequency limit value,
-wherein preferably the test frequency is varied over a plurality of frequency ranges and the frequency steps of the different frequency ranges are set different from each other, wherein preferably each frequency step is set individually to a predetermined percentage share greater than the respective lower frequency value of the associated frequency range.
10. A wind energy system, namely a wind energy installation or a wind farm, for detecting low-frequency oscillations, in particular subsynchronous resonances, in a supply grid having a grid voltage with a nominal grid frequency, and comprising:
recording means for recording at least one measurement sequence of grid variables, in particular grid voltage, feed current or grid frequency, over a measurement period for performing a frequency analysis, wherein the measurement sequence has a plurality of measurement points,
-a multiplication unit for multiplying the measurement sequence with a sinusoidal test function of simultaneous correlation in the same measurement time period, wherein
-the test function is characterized by a test frequency and a test angle as a phase angle, and
-multiplying the measurement sequence with the test function for each measurement point to obtain a test product for each measurement point,
-an adding unit for adding the test products to a product sum taking into account the sign of the test products, and
-evaluation means for evaluating, from the product sum: whether the measurement sequence has a low frequency oscillation having:
-a frequency in the range of the test frequency, and
-a phase angle within the test angle range.
11. The wind energy system of claim 10, characterized in that the wind energy system is configured to carry out the method according to any one of claims 1 to 9, wherein the wind energy system preferably has a process computer on which the method is carried out, at least a part of it.
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| DE102018120768.9A DE102018120768A1 (en) | 2018-08-24 | 2018-08-24 | Wind turbine and method for detecting low-frequency vibrations in an electrical supply network |
| DE102018120768.9 | 2018-08-24 | ||
| PCT/EP2019/072601 WO2020039077A1 (en) | 2018-08-24 | 2019-08-23 | Wind turbine and method for detecting low-frequency vibrations in an electric supply network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112602248A true CN112602248A (en) | 2021-04-02 |
Family
ID=67742438
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201980055661.5A Pending CN112602248A (en) | 2018-08-24 | 2019-08-23 | Wind energy installation and method for detecting low-frequency oscillations in an electrical supply network |
Country Status (6)
| Country | Link |
|---|---|
| US (1) | US11899050B2 (en) |
| EP (1) | EP3841650A1 (en) |
| CN (1) | CN112602248A (en) |
| CA (1) | CA3109200C (en) |
| DE (1) | DE102018120768A1 (en) |
| WO (1) | WO2020039077A1 (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111679125B (en) * | 2020-06-04 | 2021-07-20 | 北京交通大学 | Method and device for identifying oscillation of power system |
| EP4170848A1 (en) | 2021-10-20 | 2023-04-26 | Wobben Properties GmbH | Method for damping low-frequency vibrations in an electrical supply network |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA1242484A (en) * | 1985-11-20 | 1988-09-27 | Ecole Polytechnique | Method and system for measuring the amplitude and phase angle of harmonics in a periodic signal |
| US5677852A (en) * | 1992-09-15 | 1997-10-14 | Asea Brown Boveri Ab | Method and arrangement for detecting and damping transcient oscillations at or near a natural resonant frequency in a power transmission system |
| US20070241759A1 (en) * | 2006-04-07 | 2007-10-18 | Michael Lamar Williams | Method for measuring stability margin at a node of a polyphase power grid |
| RU2442180C1 (en) * | 2010-10-07 | 2012-02-10 | Государственное образовательное учреждение высшего профессионального образования "Северо-Кавказский государственный технический университет" | Method for determining the parameters harmonic for non-sinusoidal electrical signal |
| CN104820164A (en) * | 2015-04-14 | 2015-08-05 | 华南理工大学 | Method for positioning low-frequency oscillation disturbance source of electric power system |
| CN105472382A (en) * | 2014-09-29 | 2016-04-06 | 三星电子株式会社 | Error detector and method for detecting error of an oscillator |
| CN106841778A (en) * | 2016-12-28 | 2017-06-13 | 国电南瑞科技股份有限公司 | The processing method of the subsynchronous and supersynchronous harmonic parameters realized based on PMU |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4031462A (en) * | 1975-07-07 | 1977-06-21 | Motorola, Inc. | Frequency spectrum analyzer |
| DE3733555A1 (en) * | 1987-10-03 | 1989-04-13 | Hartmann & Braun Ag | METHOD FOR MEASURING THE AMPLITUDE OF A PERIODIC TIME-DEPENDENT ELECTRICAL SIGNAL G (T) IN A SIGNAL MIXTURE U (T) |
| US10983150B2 (en) * | 2017-08-28 | 2021-04-20 | General Electric Technology Gmbh | Systems and methods for detecting and evaluating oscillations in an electrical power grid |
| US11637428B2 (en) * | 2020-09-29 | 2023-04-25 | Mitsubishi Electric Power Products, Inc. | Selective power oscillation damping control system with damping ratio dynamic gain and method for the same |
-
2018
- 2018-08-24 DE DE102018120768.9A patent/DE102018120768A1/en active Pending
-
2019
- 2019-08-23 CN CN201980055661.5A patent/CN112602248A/en active Pending
- 2019-08-23 EP EP19758725.6A patent/EP3841650A1/en active Pending
- 2019-08-23 CA CA3109200A patent/CA3109200C/en active Active
- 2019-08-23 US US17/270,820 patent/US11899050B2/en active Active
- 2019-08-23 WO PCT/EP2019/072601 patent/WO2020039077A1/en unknown
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA1242484A (en) * | 1985-11-20 | 1988-09-27 | Ecole Polytechnique | Method and system for measuring the amplitude and phase angle of harmonics in a periodic signal |
| US5677852A (en) * | 1992-09-15 | 1997-10-14 | Asea Brown Boveri Ab | Method and arrangement for detecting and damping transcient oscillations at or near a natural resonant frequency in a power transmission system |
| US20070241759A1 (en) * | 2006-04-07 | 2007-10-18 | Michael Lamar Williams | Method for measuring stability margin at a node of a polyphase power grid |
| RU2442180C1 (en) * | 2010-10-07 | 2012-02-10 | Государственное образовательное учреждение высшего профессионального образования "Северо-Кавказский государственный технический университет" | Method for determining the parameters harmonic for non-sinusoidal electrical signal |
| CN105472382A (en) * | 2014-09-29 | 2016-04-06 | 三星电子株式会社 | Error detector and method for detecting error of an oscillator |
| CN104820164A (en) * | 2015-04-14 | 2015-08-05 | 华南理工大学 | Method for positioning low-frequency oscillation disturbance source of electric power system |
| CN106841778A (en) * | 2016-12-28 | 2017-06-13 | 国电南瑞科技股份有限公司 | The processing method of the subsynchronous and supersynchronous harmonic parameters realized based on PMU |
Non-Patent Citations (2)
| Title |
|---|
| E.JACOBSEN, ET AL: "The sliding DFT", IEEE SIGNAL PROCESSING MAGAZINE, vol. 20, no. 2, pages 74 - 80 * |
| 竺炜: "分段傅里叶神经网络的低频振荡模式识别方法", 《电力系统保护与控制》, vol. 40, no. 15, pages 40 - 45 * |
Also Published As
| Publication number | Publication date |
|---|---|
| EP3841650A1 (en) | 2021-06-30 |
| CA3109200A1 (en) | 2020-02-27 |
| US20210341523A1 (en) | 2021-11-04 |
| US11899050B2 (en) | 2024-02-13 |
| WO2020039077A1 (en) | 2020-02-27 |
| DE102018120768A1 (en) | 2020-02-27 |
| CA3109200C (en) | 2023-08-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11545832B2 (en) | Method for controlling a wind farm in order to damp sub synchronous oscillations | |
| Ibrahim et al. | An effective approach for rotor electrical asymmetry detection in wind turbine DFIGs | |
| Yang et al. | Interpolated DFT-based identification of sub-synchronous oscillation parameters using synchrophasor data | |
| Soliman et al. | Capacitance estimation algorithm based on DC-link voltage harmonics using artificial neural network in three-phase motor drive systems | |
| JP6945371B2 (en) | Diagnostic equipment for rotating machine systems, power converters, rotating machine systems, and diagnostic methods for rotating machine systems. | |
| KR20180032628A (en) | Method and apparatus for detecting voltage in a supply network | |
| CN112602248A (en) | Wind energy installation and method for detecting low-frequency oscillations in an electrical supply network | |
| CN112385109A (en) | Method for detecting low-frequency oscillations and detection device therefor | |
| CN109061297A (en) | One kind time/supersynchronous oscillation on-line monitoring method and device | |
| KR101325473B1 (en) | A motor monitoring system | |
| US11658490B2 (en) | Wind energy system and method for identifying low-frequency oscillations in an electrical supply network | |
| JP6430234B2 (en) | Vibration analysis apparatus and program for rotating machine | |
| Chen et al. | Low cost Arduino DAQ development and implementation on an Android app for power frequency measurement | |
| KR101800953B1 (en) | Energy storage system and method for controlling power frequency | |
| JP6423757B2 (en) | Frequency detection device, frequency detection method, and inverter device using detected frequency | |
| D'Apuzzo et al. | A time-domain approach for the analysis of nonstationary signals in power systems | |
| US20210208186A1 (en) | Wind turbine and method for detecting low-frequency oscillations in an electrical supply grid | |
| Junput et al. | Optimal technique for total harmonic distortion detection and estimation for smart meter | |
| Moriano et al. | A novel approach to the grid inductance estimation based on second order generalized integrators | |
| CN111679124B (en) | Oscillation identification method and device for power system | |
| Wu et al. | Design of TCSC damping controller for SSO suppression based on TLS-ESPRIT | |
| Mahmud et al. | A Smart Monitoring System for Rotor Bar Fault Detection in Induction Motors | |
| Florencias-Oliveros et al. | Voltage Supply Frequency Uncertainty influence on Power Quality index: A qualitative analysis of Higher-Order Statistics 2D trajectories | |
| Van et al. | SVR-based flicker estimation for wind power systems | |
| Wang et al. | Analysis of transients in a micro-grid using wavelet transformation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |